J. Org. C h e m . 1991,56,2197-2202 41) for elution to yield 35 mg (8%) of (t)-(R)-13[[a]@H, t21O ( c 0.3,CHC13, optically pure)] and 13 mg (2%) of l-fluoro-l(phenylseleno)-2,2-diphenylcyclopropane[mp 91-3 OC; [aI2'H# +221° (C 0.12,CHC13); 'H N M R (CDCl3)6 2.12 (1H, dd, J = 7.2, 8.4),2.16 (1H, dd, J = 7.2,14.4),7.2-7.35(11H, m), 7.4-7.55 (4 H, m)l. Anal. Calcd for CZlHl7FSe:C, 68.66;H, 4.66. Found C, 68.49; H, 4.68. Phenyl (+)-(R)-l-fluoro-2,2-diphenylcyclopropaneselenocarboxylatewas isolated in 48% yield (0.38 9): mp 114-5",[.IU& +339O (c 1.2,CHCI,); 'H NMR (CDCl,) 6 2.25 (1H, dd, J = 6.6, 18.3),2.47 (1H, dd, J = 6.6,9.3),7.1-7.5 (15H, m); IR (Nujol) 1710,1500,9.55 cm-'.
2197
Anal. Calcd for Cz2H1,FOSe: C, 66.84;H, 4.33. Found: C, 66.73;H, 4.38. E. A 10-mL benzene solution of ester 3 (2mmol prepared as in D above) and 1 mL of thiophenol were refluxed for 0.5 h. The usual workup procedure afforded phenyl disulfide (0.35g) (mp 57-8 O C ; (+)-(R)-13(0.16g, 38%) [[aIz4Q +21.0° (c 1.5,CHCl,, optically pure)];and phenyl (+)-(R)-l-fluoro-2,2-diphenylcyclopropanethiocarboxylate (0.115 g, 18%) [mp '103-4 O C (from hexane); [aI2'Hg +402O (c 0.5, CHCI,); lH NMR (CDCl,) 6 2.21 (1 H, dd, J = 6.9,18.3),2.50 (1H, dd, J = 6.9,9.6),7.1-7.4 (13 H, m), 7.52 (2 H, d, J = 8); IR (Nujol) 1700, 1500, 970 cm-'. Anal. Calcd for CBH17FOS:C, 75.84;H, 4.92. Found C, 75.82; H, 4.97.
Absolute Rate Constants for the @-ScissionReaction of the 1-Phenyl-2-phenoxypropylRadical: A Model for Radical Reactions of Lignin' S. Thomas Autrey, Mikhail S.Alnajjar, David A. Nelson, and James A. Franz* Battelle, Pacific Northwest Laboratory, Richland, Washington 99352
Received June 25,1990
Absolute rate expressions for @-scissionof the phenoxy radical from the 1-phenyl-2-phenoxypropylradical, forming cis- and trans-@-methylstyrene,were determined by competition of @-scission(k,) with abstraction of hydrogen from trimethylstannane (k.b). Relative rates (k,/kab) were converted to absolute rates (ks)by using a rate expression determined for abstraction of hydrogen atom from tributylstannane by the phenylethyl radical: log [k+/(M-l s-l)] = (9.31f 0.30)- (7.11f 0.49)/8,where 0 = 2.303RT kcal/mol (errors are 2u). The resulting expressions for @-scissionare log (kAb,/s-l) = (13.45f 0.26)- (16.94f 0.52)/8and log (ks,&/s-l) = (13.41f 0.3)- (19.3f 0.75)/8. The basis rate expression for abstraction of hydrogen from tributylstannaneby the phenylethyl radical was determined in a competition of abstraction (kab) with self-termination(kJ,using the Smoluchowski expression for self-reaction of phenylethyl radical: log [2k,/(M-ld)]= 11.93 - 3.112/8. Combiningthe Arrhenius parameters with the enthalpy change for 8-scission leads to activation barriers for addition of phenoxy radical to trans- and cis-b-methylstyreneof 5.2and 6.6 kcal/mol, respectively.
Introduction A predominate structural cross-link in the macromolecular network of lignin is the aryl @-arylether linkage, -(ArOCCAr)-.2 Solid-state NMR studies suggest that alkyl aryl ether linkages may be present in modest extent in low-rank coalsS3 A detailed understanding of the free-radical4 and radical cation6 chemistry of lignin is necessary for the design of new processes of pulp preparation,6 the understanding of pathways of hydrothermal conversion of biomass and coal to useful products,6 and understanding the process of coalificationof lignocellulosic s t r ~ c t u r e . ~The thermal decomposition of model com(1) This work was supported by the Office of Basic Energy Sciences,
US. Department of Energy (DOE),.under contract DE-AC06-76RLO1830 with Battelle Memorial Institute, which operates the Pacific Northwest Laboratory for DOE. (2) Pearl, I. A. The Chemistry of Lignin; Marcel Dekker: New York, 1967. (3) Lignite may contain up to 12% oxygen-substituted aliphatic carbon: Solum, M. s.; Pugmire, R. J.; Grant, D. M. Energy Fuels 1989,3, 187-193. (4) Simkovic,I. J. Macromol. Sci., Reo. Macromol. Chem. Phys. 1986, C26,67. (5) DiCoeimo, R.; Szabo, H.-C. J. Org. Chem. 1988,53, 1673. (6) Poutama, M. L. A Review of Thermolysis of Model Compounds Relevant to Processing of Coal. Oak Ridge National Laboratory Technical Document No. ORNL/TM-10637; Nov 1987. Available from the National Technical Information Service, U.S.Department of Commerce, 5285 Port Royal Road, Springfield, VA 22161. (7) (a) Winans, R. E.; Hayatau, R.; Squires, T. G.; Carrado, K. A.; Both, R. E. Prep.-Am. Chem. SOC., Diu. Pet. Chem. 1990,35(2),423. (b) McMillen, D. F.; Malhotra, R. Prepr.-Am. Chem. SOC.,Diu. Pet. Chem. 1990, 35(2), 430.
pounds6 and polymers7 containing known structural units of lignin or proposed structural links in coal under hydroliquefactionconditions is a valuable exercise since dired observation of structurally distinct reactions remains nearly impossible for coal and difficult for lignin. Under ideal circumstances, the global kinetics and kinetic reaction order of thermal decomposition reactions of model compounds containing linkages of relevance to lignin and coal structure can be reduced to the contributing individual stepwise rate constants. The careful studies by Poutsma and Dyer8 and Gilbert and Gajewskigof the homogeneous thermal decomposition of 1,n-diphenylalkanes (n = 2-4) and the studies of Buchanan and co-workers of heterogeneous decomposition of similar structures bonded to silica surfaces'" provide examples of successful reduction of global rates to individual contributing rates. However, even the early stages of thermal decompositions of nominally simple systems may involve multiple initiation and propagation steps and early participation of secondary reactions. For these cases, design of experiments to directly determine individual reaction steps is desirable. A recent model compound studyg examined radical chain decomposition pathways for cleavage of the C-0 bond in phenyl 2-phenylethyl ether, as a model of reactions of similar structures presumably in low-rank coals. A free(8) Poutama, M. L.; Dyer, C. W. J. Org. Chem. 1982,47, 4903. (9) J. J. J . Ora Chem. 1982. 47. 4899. ~. , Gilbert. -. . -. . K. E -. ,:Gaiewski. -(10) Buchanan,A. C.,*III;-Dunstan,T. fi J.; Douglas, E. d;Poutama, M. L. J . Am. Chem. SOC.1986,108, 7703.
.--
OO22-3263/91/1956-2197$02.50/0 0 1991 American Chemical Society
2198 J. Org. Chem., Vol. 56, No. 6,1991
Autrey et al.
radical chain decomposition pathway was suggested for the thermal decomposition of phenyl 2-phenylethylether (eqs 1-4). The decomposition of this ether is initiated in the early stage of decomposition by unimolecular cleavage of the ether to form phenoxy1 and phenylethyl radicals (eqs 1 and 2). Chain-propagation steps include abstraction of the benzylic hydrogen of the ether and j3-scission of the resulting 1-phenyl-2-phenoxyethylradical to form styrene and phenoxy radical (eq 3), followed by termination reactions (eq 4). A global activation barrier of 50.3 kcal/mol PhCHzCHzOPh PhCHzCHzOPh + PhO'
-
PhCH(')CH20Ph 2Ph0' PhO'
-C
PhCH2CH2' + 'OPh (1) PhCH(')CHZOPh + PhOH (2)
-
PhCH=CH2
+ PhCH(')CH20Ph-. 1
+ 'OPh
(3)
nonradical products (4)
was reported for the decomposition of the ether, compared to 65 kcal/mol for unimolecular scission of the C-O bond (eq l),reflecting a chain process. A barrier for /3-scission (eq 3) was reported to be 28 kcal/mol? although details of how this value was obtained were not given. This value matches the barrier for @-scissionof 1,3-diphenylpropyl radical: for which Poutsma and Dyer estimate log (kS/s-l) = 14.8 - 28.3/8, 8 = 2.3RT kcal/mol:
-
PhCH(*)CHZCH2Ph
PhCH=CHZ
+ PhCH2'
(5)
The value of 28 kcal/mol for eq 3 is in significant disagreement with a recent investigation of substituent effects in the &scission of PhCH(')CH(CH,)OPh in competition with hydrogen abstraction from trimethylstannane." That study estimated a barrier of no more than 19 kcal/mol for the j3-scission reaction of the structurally similar radical. The discrepancy may reflect the incursion of a variety of second-generation reactions in addition to those of eqs 1-4, making analysis of initial rates of reaction and the overall reaction order difficult and ambiguous. Thus, to provide an accurate value for the 0-scission reaction, and to provide a kinetic basis for prediction of radical-based degradation pathways for lignin or low-rank coal structure, we have determined an absolute rate expression for the &scission of the 1-phenyl-2-phenoxypropyl radical. Arrhenius parameters for scission of this model of the ArCCOAr linkage in lignin by a free-radical pathway and estimates for the reverse reaction, addition of phenoxy radical to P-methylstyrene, are presented.
Results a n d Discussion Abstraction of Hydrogen from Tributylstannane by Phenylethyl Radical. Although a rate expression for abstraction of hydrogen atom from BuBSnHby benzyl radical has been published,12rate constants for the reaction of alkyl-substituted benzylic radicals with Bu3SnH or Me3SnH, suitable for use as basis rate constants for conversion of relative rates of &scission vs abstraction (ks/kabs) to absolute rates (kB),have not appeared. Thus, an Arrhenius expression for abstraction of hydrogen by the phenylethyl radical from Bu3SnH (kabs) was determined by a competition of self-termination of the phenylethyl radical with abstraction. The self-termination rates were in turn calculated by using the Smoluchowski equation (eq (11) Suleman, N. K.; Nelson, D. J. Org. Chem. 1989,54, 503. (12) Franz, J. A.; Suleman, N. K.; Alnajjar, M. S. J. Org. Chem. 1986,
51, 19.
Table I. Arrhenius Parameters and Rate Constants for Abstraction of Hydrogen by Alkyl and Benzylic Radicals from Tributylstannane" k(25 "C) solvent radical E. x Id log A 1.1 benzene phenylethyP 9.24 i 0.30 7.11 i 0.49 benzyl' 8.65 f 0.17 5.58 f 0.24 3.6 C-C~HI~ 1050 isooctane 9.39 i 0.28 3.23 t 0.34 methyld 146 boddine isopropyld 8.71 f 0.37 3.47 f 0.49 "Errors are 20. bThis work. 'Reference 12. dReference 20.
6), which predicts rates of self-reaction of small, carboncentered radicals in nonassociating solvents of low viscosity with typically less than 15% error:13-16 2kt = (8~/1000)apDA$V
(6)
By analogy with the benzyl radical and other small hydrocarbon radicals, the phenylethyl radical will exhibit a spin statistical factor (a) near 1/4, the fraction of singlet radical pairs, reflecting inefficient intersystem crossing between the singlet and triplet states'during the lifetime of the solvent cage. Values of u near 1/4 have been demonstrated for tert-butyl, benzyl, isopropyl, and oxygensubstituted carbon-centered radicals, 2-hydroxyprop-2-yl and hydroxymethyl, in alkane, benzene, acetonitrile, methanol, and tetraethoxysilane solvents, among others.13J7 The diffusion coefficient of eq 6 is given by D,q/T = k/67rrAf, where f is the microfriction factor of Spemol and Wirtz.15 To check the accuracy of the Spernol-Wirtz diffusion coefficients, the diffusion coefficients of ethylbenzene, the model for the phenylethyl radical, in benzene at 25 and 60 "C were determined and found to agree within 15% of predicted values. The parameter p is the average of the LeBas,14van der W a a W and Spern~l-Wirtz'~ reaction diameters, and N is Avogrado's number.', The Smoluchowski expression for total self-termination rate (disproportionation plus combination) for the phenylethyl radical in benzene is given by In [2kt/(M-' s-l)] = 27.49 - 3112/RT,18 or log [2kt/(M-' s-l)] = 11.93 - 3.112/8. To determine the Arrhenius expression for abstraction by phenylethyl radical, we irradiated 2-phenylpropiophenone in benzene and Bu3SnH. Low conversion of the ketone photoprecursor and hydride donor produces phenylethyl radicals at constant concentration (eqs 7-10). Abstraction of hydrogen by the phenylethyl radical from the stannane to produce ethylbenzene competes with combination of the phenylethyl radical to yield dimers (l), (13) Schuh, H.-H.; Fischer, H. Helu. Chim. Acta 1978,61,2130. (14) Ghai, R. K.; Dullien, F. A. L. J . Phys. Chem. 1974, 78, 2283. (15) Spernol, A.; Wirtz, K. Z. Naturforsch. 1953,8a, 622-532. (16) Edward, J. T. J. Chem. Educ. 1970,47(4), 261-270. (17) Saltiel, J.; Atwater, B. W. Spin Statistical Factora in DiffwionControlled Reactions. In Adoances in Photochemistry;Volman, D. H., Hammond, G. S., Gollnick, K., Eds.;Wiley-Interscience: New York, 1988, VOI. 14, pp 1-90. (IS)The Smoluchoweki equation for self-reaction of phenylethyl radicals in benzene waa calculated by wing the following date: radical model, ethylbenzene, mp 178.2 K, bp 409.3 K, p 6.33 X 10% cm (averageof GD (ref 14), van der Waals (ref 161, and Spernol-Wirtz (ref 15) reaction diameters, see ref 13);solvent, benzene MW 78.11, mp 278.7 K, bp 353.3 K Andrade viscosity of benzene, In (7,CP) = -4.117 + 2093.7/Rc ethylbenzene density, dens(p/mL) = 1.134-9.0913X l0-l X T(K); benzene dens(g/mL) = 1.192 - 1.059 X IO-* X T(K). The microfriction factor of Spemol and Wirtz ia given by f = (0.16 +0.4rA/rB)(0.9+ 0.4TAr- 0.25T~3. Reduced temperatures are given by Txr= (T- Tx?/(Txb - Tx?,where Tx'and TXbare freezing and boiling points of species X = phenylethyl (A) or benzene (B).The radii in the microfraction factor term are given by rx = (3Vx(x)/4rN)'18,where x = 0.74, the volume fraction for cubic closest packed spheres. The resulting expression for self-terminationof phenylethyl radical is In (2ktlM-I s-') = 27.485 - 3112/RT, or log (2kt/M-' s-') 11.93 - 3.112/0.
-
Model for Radical Reactions of Lignin
J. Org. Chem., Vol. 56, No.6, 1991 2199 Table 11. Arrheniur Exprerrions for @-ScirrionReactionsa
-.
PhCH(')CH(CH&OPh PhCH==CHCHB (cis + trans) + PhO(') E,, kcal/mol k(298K,s-')
log (Ala-')
reaction
* 0.26 13.41 f 0.3
*
trans-@-scission 13.45
16.94 0.52
cis-@-scission
19.3 & 0.75
10.7 0.18
Errors are 2u,
1.2
1.9
1.4
i.8
1.6
1.7
l/RT x 1000
Figure 1. Arrhenius plot for abstraction of hydrogen by the phenylethyl radical from tributylstannanein benzene, 80-160 O C .
meso- and d,l-2,3-diphenylbutane, and disproportionation to yield both PhCH=CH2and PhCH2CHS. To determine PhCOCH(CH3)Ph-!!L PhCO'
+ PhCH(')CH,
(7)
PhCH(')CH3 + Bu3SnH -kPhCH2CH3
--
(8) 2PhCH(')CH3 P h C H e H 2 + PhCH2CH3 (9) 2PhCH(')CH3 PhCH(CHJCH(CH3)Ph (10) 1, meso + d,l the fraction of the total yield of ethylbenzene produced by abstraction from tributylstannane (kab), the total yield of ethylbenzenemust be corrected for ethylbenzeneformed during the disproportionation step (eq 9). The temperature dependence of the ratio of disproportionation to combination,kg/klo,for phenylethyl radical in benzene has been determined in a careful study: kg/klo = exp(1.47/RT + 8.43/R).l9 Thus, the yield of ethylbenzene produced in the disproportionation step can be calculated at each temperature from the yield of meso and d,l dimers (1). The rate of abstraction, knb, is given by eq 11,and the rate of total termination, kt = kg + klo, calculated from the Smoluchowski equation, is given in eq 12. Under cond[PhCH2CH3(abs)/dt = knb,[Bu3SnH][PhCH(')CH,] (11)
d([ll + (kg/k,o)[ll)/dt = kJPhCH(')CH3I2 (12) ditions of short (
